| *** mrda has quit IRC | 00:07 | |
| *** tosky has quit IRC | 00:12 | |
| mgariepy | thanks jonher i already have this patch i did cherrypicked the last bump ;) | 00:24 |
|---|---|---|
| *** ianychoi__ is now known as ianychoi | 00:34 | |
| *** aedan has quit IRC | 00:55 | |
| *** macz_ has quit IRC | 01:16 | |
| *** macz_ has joined #openstack-ansible | 01:17 | |
| *** macz_ has quit IRC | 01:22 | |
| *** jhesketh_ is now known as jhesketh | 01:49 | |
| *** poopcat1 is now known as poopcat | 03:05 | |
| *** maharg101 has joined #openstack-ansible | 03:37 | |
| *** rh-jlabarre has quit IRC | 03:42 | |
| *** maharg101 has quit IRC | 03:43 | |
| *** dmsimard has quit IRC | 04:08 | |
| *** dmsimard has joined #openstack-ansible | 04:09 | |
| *** evrardjp_ has quit IRC | 05:33 | |
| *** evrardjp has joined #openstack-ansible | 05:33 | |
| *** maharg101 has joined #openstack-ansible | 05:39 | |
| *** maharg101 has quit IRC | 05:44 | |
| *** miloa has joined #openstack-ansible | 07:16 | |
| *** miloa has quit IRC | 07:30 | |
| openstackgerrit | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_adjutant master: Install mysql client libraries https://review.opendev.org/c/openstack/openstack-ansible-os_adjutant/+/777607 | 07:31 |
| *** masterpe has joined #openstack-ansible | 07:42 | |
| masterpe | Doing a upgrade to ussuri with our own apt mirrors I notice that, openstack_hosts/vars/ubuntu-18.04.yml uses apt_repo_url and openstack_hosts/vars/ubuntu-20.04.yml uca_apt_repo_url as a variable and why doesn't uses the same variable? | 07:42 |
| *** andrewbonney has joined #openstack-ansible | 08:24 | |
| openstackgerrit | Andrew Bonney proposed openstack/openstack-ansible-os_octavia master: Add variable to override keypair setup python interpreter https://review.opendev.org/c/openstack/openstack-ansible-os_octavia/+/772559 | 08:25 |
| *** luksky has joined #openstack-ansible | 08:26 | |
| *** maharg101 has joined #openstack-ansible | 08:35 | |
| *** pto has joined #openstack-ansible | 08:41 | |
| openstackgerrit | Dmitriy Rabotyagov proposed openstack/openstack-ansible stable/victoria: Bump SHAs for stable/victoria https://review.opendev.org/c/openstack/openstack-ansible/+/775521 | 08:51 |
| *** shyamb has joined #openstack-ansible | 09:13 | |
| *** shyamb has quit IRC | 09:14 | |
| *** shyamb has joined #openstack-ansible | 09:14 | |
| openstackgerrit | Andrew Bonney proposed openstack/openstack-ansible master: Update setuptools to match version in upper-constraints https://review.opendev.org/c/openstack/openstack-ansible/+/777730 | 09:33 |
| openstackgerrit | Andrew Bonney proposed openstack/openstack-ansible-os_neutron master: Combined patch to unblock CI https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/775239 | 09:34 |
| *** pto has quit IRC | 09:34 | |
| *** mubix has quit IRC | 09:37 | |
| *** mubix has joined #openstack-ansible | 09:37 | |
| openstackgerrit | Andrew Bonney proposed openstack/openstack-ansible-os_octavia master: Move octaiva pip packages from constraints to requirements https://review.opendev.org/c/openstack/openstack-ansible-os_octavia/+/772293 | 09:43 |
| jrosser | masterpe: that looks like an error in the vars/ubuntu-20.04.yml file, though if you note just above that line _package_repos: [] is set for focal, so it's not used anyway on 20.04 | 09:53 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-openstack_hosts master: Correct the UCA repo location variable for ubuntu focal https://review.opendev.org/c/openstack/openstack-ansible-openstack_hosts/+/777734 | 09:57 |
| *** klamath_atx has joined #openstack-ansible | 10:02 | |
| *** gshippey has joined #openstack-ansible | 10:13 | |
| *** shyamb has quit IRC | 10:14 | |
| *** mugsie_ is now known as mugsie | 10:16 | |
| *** shyamb has joined #openstack-ansible | 10:36 | |
| *** klamath_atx has quit IRC | 11:09 | |
| *** klamath_atx has joined #openstack-ansible | 11:21 | |
| *** tosky has joined #openstack-ansible | 11:25 | |
| *** ajg20 has quit IRC | 11:26 | |
| openstackgerrit | Merged openstack/openstack-ansible-openstack_hosts master: Use ansible_facts[] instead of fact variables https://review.opendev.org/c/openstack/openstack-ansible-openstack_hosts/+/776960 | 11:49 |
| *** shyamb has quit IRC | 12:10 | |
| *** shyamb has joined #openstack-ansible | 12:26 | |
| *** shyamb has quit IRC | 12:33 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_adjutant master: Use venv_tag https://review.opendev.org/c/openstack/openstack-ansible-os_adjutant/+/777597 | 12:47 |
| openstackgerrit | Merged openstack/openstack-ansible-os_adjutant master: Use global service variables https://review.opendev.org/c/openstack/openstack-ansible-os_adjutant/+/777596 | 12:47 |
| mgariepy | good morning everyone. | 13:20 |
| *** zigo has joined #openstack-ansible | 13:20 | |
| mgariepy | jrosser, so I disabled 2 of the repo container in haproxy. i still have the same contaienr not deployning. | 13:25 |
| openstackgerrit | Merged openstack/openstack-ansible master: Workaround nova bug https://review.opendev.org/c/openstack/openstack-ansible/+/777601 | 13:40 |
| mgariepy | http://paste.openstack.org/show/803037/ | 14:01 |
| mgariepy | why the : Skipping link: unsupported archive format: .04-x86_64: http://10.41.15.2:8181/os-releases/22.0.2.dev8/ubuntu-20.04-x86_64 (from -f) | 14:01 |
| jrosser | that is very odd indeed | 14:06 |
| mgariepy | i did wipe and recreate the container | 14:08 |
| jrosser | it’s like pip is parsing the url and not liking the dot | 14:10 |
| mgariepy | le me regen a cleanish log i'll upload it whole.. | 14:11 |
| *** jpvlsmv has quit IRC | 14:16 | |
| openstackgerrit | Georgina Shippey proposed openstack/openstack-ansible master: [WIP] Use the Mitogen connection method in OSA https://review.opendev.org/c/openstack/openstack-ansible/+/591236 | 14:17 |
| mgariepy | https://paste.ubuntu.com/p/MTyPWZnN86/ | 14:17 |
| mgariepy | 2k lines of pleasure. | 14:18 |
| *** jpvlsmv has joined #openstack-ansible | 14:18 | |
| openstackgerrit | Georgina Shippey proposed openstack/openstack-ansible master: [WIP] Use the Mitogen connection method in OSA https://review.opendev.org/c/openstack/openstack-ansible/+/591236 | 14:20 |
| openstackgerrit | Andrew Bonney proposed openstack/openstack-ansible-tests master: DNM: Pull constraint files from local path in CI https://review.opendev.org/c/openstack/openstack-ansible-tests/+/777753 | 14:22 |
| mgariepy | it finds the file in the repo but it seems to prefer the upstream one for ${REASON} | 14:23 |
| *** ajg20 has joined #openstack-ansible | 14:51 | |
| openstackgerrit | Andrew Bonney proposed openstack/openstack-ansible-os_neutron master: Combined patch to unblock CI https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/775239 | 14:56 |
| ajg20 | Any one have any advice to resolve this Openstack ansible AIO issue? | 15:02 |
| ajg20 | fatal: [aio1_horizon_container-1804132b -> 172.29.236.100]: FAILED! => {"changed": false, "msg": "The `lxc` module is not importable. Check the requirements."} | 15:02 |
| ajg20 | I am running | 15:03 |
| ajg20 | sudo openstack-ansible /opt/openstack-ansible/playbooks/setup-hosts.yml | 15:03 |
| mgariepy | jrosser, for some reason, the first container gets a constraint that is 1471 in size. the 2 others gets a 212 one. | 15:03 |
| mgariepy | which have the git url in so. that's probably the reason why it does with it do.. | 15:03 |
| jrosser | mgariepy: the git url should never be in the target host constraints | 15:09 |
| jrosser | only on the build host | 15:09 |
| mgariepy | i know, something is wrong here. | 15:09 |
| mgariepy | https://github.com/openstack/ansible-role-python_venv_build/blob/stable/victoria/tasks/python_venv_install.yml#L85-L98 | 15:10 |
| *** tosky has quit IRC | 15:19 | |
| *** tosky has joined #openstack-ansible | 15:20 | |
| jrosser | mgariepy: it could also have tried (and succeeded) to do the build locally on the first host rather than use the repo server | 15:32 |
| jrosser | if somehow venv_build_host did not evaluate correctly (like stale facts cache?) | 15:33 |
| *** macz_ has joined #openstack-ansible | 15:37 | |
| *** macz_ has quit IRC | 15:38 | |
| *** macz_ has joined #openstack-ansible | 15:39 | |
| mgariepy | having kids around is not great for focussing on sstuff haha | 15:46 |
| mgariepy | somehow the constraints it's getting is the sources-constraint from the repo container. | 16:39 |
| *** gyee has joined #openstack-ansible | 16:40 | |
| jrosser | mgariepy: the wheels that get built are indexed on the repo server and the constraints transferred to the target should come from here https://github.com/openstack/ansible-role-python_venv_build/blob/master/tasks/python_venv_wheel_build.yml#L159-L173 | 16:41 |
| mgariepy | http://paste.openstack.org/show/803040/ | 16:44 |
| mgariepy | this correspond to the source-constraints.txt | 16:45 |
| mgariepy | hm. is it possible that since the task is delegated only on the first host that the var is not accessible to the other hosts? | 16:47 |
| jrosser | i just checked a multinode lab here and none of the placement container venv constraints have the git url in them | 16:54 |
| mgariepy | on V ? | 16:54 |
| jrosser | yeah, and source-constraints file is only made on the host that does the build | 16:54 |
| jrosser | this all points to it not knowing to build on the repo server | 16:55 |
| *** strattao has joined #openstack-ansible | 16:55 | |
| jrosser | which is decided by this https://github.com/openstack/ansible-role-python_venv_build/blob/master/tasks/python_venv_wheel_build.yml#L24 | 16:55 |
| jrosser | perhaps throw in a debug just there to print venv_build_host | 16:56 |
| mgariepy | _constraints_file_slurp is not defined on the 2nd and 3rd nodes | 16:56 |
| mgariepy | only on the first one.. | 16:57 |
| mgariepy | ha. | 17:00 |
| mgariepy | hmm.. | 17:01 |
| mgariepy | lol. :( | 17:01 |
| mgariepy | https://github.com/openstack/ansible-role-python_venv_build/blob/master/tasks/python_venv_wheel_build.yml#L171 < this is only registered on the first container. | 17:04 |
| mgariepy | https://github.com/openstack/ansible-role-python_venv_build/blob/master/tasks/python_venv_install.yml#L105 | 17:05 |
| jrosser | but that entire block should be delegated? | 17:06 |
| mgariepy | it is on the first node. | 17:06 |
| jrosser | so i think its still a good question why that doesnt happen for the 2nd/3rd | 17:10 |
| mgariepy | running a few mode debug .. | 17:10 |
| mgariepy | arf.. lol. | 17:11 |
| mgariepy | debug msg fail. then the host are dropped lol.. | 17:12 |
| jrosser | you should print venv_build_targets as well | 17:13 |
| mgariepy | venv_build_host is set correct for all of them | 17:14 |
| mgariepy | but ghe gather bui;ld target facts is delegating to the repo server on the same host. | 17:15 |
| mgariepy | http://paste.openstack.org/show/803042/ | 17:17 |
| * jrosser looks at https://github.com/openstack/ansible-role-python_venv_build/commit/1060c112ac1d2d1816830e396e8354a2ab271044 | 17:20 | |
| jrosser | but doesnt really understand why that would do that | 17:20 |
| jrosser | i wonder if import->include has broken the very imporant side effect that run_once has | 17:22 |
| jrosser | which is to apply facts (and likley registered vars) to all play hosts even though the tasks only run against one of them | 17:22 |
| mgariepy | the build is only delegated on the first host. therefore the _constraints_file_slurp is not defined on the otehr 2 nodes http://paste.openstack.org/show/803043/ | 17:23 |
| jrosser | but run_once in main.yml is important like i say | 17:23 |
| jrosser | set_fact and register behave differently with run_once: true | 17:23 |
| *** maharg101 has quit IRC | 17:23 | |
| jrosser | try reverting the change to main.yml from the commit i linked | 17:24 |
| jrosser | i'm suspicious of that because our V upgrade was done before that code was part of a release | 17:25 |
| mgariepy | i did took the last bump in osa | 17:27 |
| mgariepy | https://review.opendev.org/c/openstack/openstack-ansible/+/775521 | 17:27 |
| mgariepy | PS 3.. | 17:28 |
| mgariepy | lol | 17:28 |
| mgariepy | reverting the patch fixes the issue. | 17:30 |
| jrosser | /o\ doh | 17:31 |
| * jrosser -W's the sha bump patch | 17:33 | |
| jrosser | yeah so run_once is a really useful construct, becasue set_fact + delegate_to + run_once: true all combined end up setting the fact on all the play hosts even though the task only ran against the delegated host | 17:34 |
| mgariepy | i isually do run on a testbed first hahah | 17:36 |
| mgariepy | not that time and i got bitten haha | 17:36 |
| mgariepy | so we revert the main.yml to import_tasks then ? | 17:39 |
| mgariepy | thanks again for your help jrosser | 17:42 |
| jrosser | we should certainly revert (and backport the revert to V) of the changes to main.yml | 17:45 |
| jrosser | no worries for helping out :) | 17:45 |
| jrosser | do you have a moment to make a patch for that? | 17:45 |
| mgariepy | sure | 17:46 |
| jrosser | awesome thanks | 17:46 |
| mgariepy | i'll do it. | 17:46 |
| mgariepy | while the remaining playbook are running :D | 17:46 |
| mgariepy | haha | 17:46 |
| mgariepy | should i revert the tag also ? | 17:49 |
| openstackgerrit | Marc Gariépy proposed openstack/ansible-role-python_venv_build master: Import_tasks instead of include_tasks https://review.opendev.org/c/openstack/ansible-role-python_venv_build/+/777798 | 17:53 |
| openstackgerrit | Marc Gariépy proposed openstack/ansible-role-python_venv_build master: Import_tasks instead of include_tasks https://review.opendev.org/c/openstack/ansible-role-python_venv_build/+/777798 | 17:56 |
| *** andrewbonney has quit IRC | 18:15 | |
| *** spatel has joined #openstack-ansible | 18:36 | |
| *** BoogieMan has joined #openstack-ansible | 19:08 | |
| BoogieMan | Hey everyone. I am trying to setup a dev environment on a hardware desktop (12 core, 32gb, nvme ssd) running fresh ubuntu 20.04 install, fully updated. I am following the guide on openstack docs, and I get through all the playbooks fine up until setup-openstack. I am consistently getting stuck at: os_keystone role, task: bootstrap admin and endpoint | 19:12 |
| BoogieMan | I let it run all night. still hangin' out here. No errors, no timeouts, no retries. Checked the logs of the keystone lxc container. I see an error about a selfed signed cert for rabbit in the keystone service journal - could be related? or red herring? | 19:13 |
| BoogieMan | Restarting the playbook hangs at the same point | 19:15 |
| *** maharg101 has joined #openstack-ansible | 19:19 | |
| *** maharg101 has quit IRC | 19:24 | |
| *** jpvlsmv has quit IRC | 19:25 | |
| mgariepy | AttributeError: module 'os_brick.initiator' has no attribute 'DRBD' | 19:29 |
| mgariepy | that's fun haha | 19:29 |
| *** jpvlsmv has joined #openstack-ansible | 19:35 | |
| *** juanoterocas has joined #openstack-ansible | 20:08 | |
| mgariepy | the override for a repo pinned to a tag in my user_var was working as expected, just that i was pointing to a Usuri repo and not a victoria one.. | 20:15 |
| *** luksky has quit IRC | 20:19 | |
| *** luksky has joined #openstack-ansible | 20:19 | |
| jrosser | BoogieMan: it’s helpful if you can paste some of the output up to the point it stops to paste.openstack.org | 20:19 |
| BoogieMan | Sure. I am doing a fresh run now. I will paste once it gets stuck. | 20:20 |
| jrosser | are you using the all-in-one? | 20:22 |
| BoogieMan | Correct | 20:22 |
| jrosser | and which branch did you use? | 20:22 |
| BoogieMan | I've tried a couple. latest 22 and a couple of the latest 21.x. let me see what it is now | 20:23 |
| BoogieMan | original runs last night were 21.2.3 and 21.2.2 after .3 failed. now it's 22 rc1 | 20:24 |
| BoogieMan | following this guide: https://docs.openstack.org/openstack-ansible/victoria/user/aio/quickstart.html | 20:26 |
| jrosser | i woudnt recommend just switching tags to try to "find something that works" | 20:26 |
| jrosser | there is an almost mechanised release every couple of weeks | 20:27 |
| jrosser | also anything with an 'rc' is a release candidate, i.e a pre-release for a particular branch | 20:28 |
| jrosser | so i would recommend using 22.0.1 as the latest tag for victoria | 20:29 |
| BoogieMan | Yeah I am aware. I would not have tried it, if not for issues with a more stable version | 20:29 |
| BoogieMan | ok 22.0.1, thats def older than I was trying. I will try it, thanks! | 20:29 |
| jrosser | well, it shouldnt be like that. the AIO config is what gets run as CI jobs dozens of times a day so it's well tested | 20:30 |
| jrosser | if theres something broken it's better we try to understand that and get it fixed | 20:30 |
| BoogieMan | my mistake, I was using latest doc not victoria. that is why I was getting much newer tags. I will try with stable/victoria now | 20:33 |
| *** devtolu1__ has joined #openstack-ansible | 20:36 | |
| *** juanoterocas has quit IRC | 20:38 | |
| *** juanoterocas has joined #openstack-ansible | 20:40 | |
| *** devtolu1__ has quit IRC | 20:43 | |
| *** spatel has quit IRC | 21:10 | |
| *** gshippey has quit IRC | 21:52 | |
| *** ajg20 has quit IRC | 22:06 | |
| *** luksky has quit IRC | 22:33 | |
| *** luksky has joined #openstack-ansible | 22:45 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_horizon master: Use ansible_facts[] instead of fact variables https://review.opendev.org/c/openstack/openstack-ansible-os_horizon/+/777651 | 22:53 |
| *** juanoterocas has quit IRC | 22:59 | |
| *** luksky has quit IRC | 23:33 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!